On 21 March 2017 at 07:40, Florian Schmaus <f...@geekplace.eu> wrote: > On 20.03.2017 23:31, Dave Cridland wrote: >> Ah, but Kev noted to me that this is using namespaced attributes. >> >> We have, traditionally, avoided these, on the basis that nobody (well, >> almost nobody) understands them. >> >> I don't think these are actually required here; a child element will >> work just as well. Can we do that instead? > > While I am aware of the "problem" of prefixes, it was a deliberate > decision to use them in ISR-SASL2. > > First, I don't think they introduce any interoperability issues here, > since their usage is negotiated by obtaining the ISR key, and they are > only used between the two immediate endpoints of the XMPP stream. > Secondly, I think it is more elegant to use the them in this particular > case. I also was unsure if *not* using them, i.e., putting a child > element within XEP-0198's <enabled/>, would violate XEP-0198's schema > (although I think it does not). >
I don't think you need to conform to the schema of a base specification if both sides have agreed to deviate by negotiation, although we almost always use a distinct namespace, and tacitly assume our schemas implicitly say any extension elements in another namespace are permitted anywhere. > I'd like to continue using them for ISR-SASL2 because I don't think they > cause any issues. I think Evgeny's message, and my response, suggests otherwise. I agree they're a sensible option here from a design perspective, but I don't think we lose anything with a child element. Dave. _______________________________________________ Standards mailing list Info: https://mail.jabber.org/mailman/listinfo/standards Unsubscribe: standards-unsubscr...@xmpp.org _______________________________________________